Figure 1.
Surface marker expression profile.
Representative diagrams of flow cytometry analysis of CD44 (FITC) and CD133 (PE) expression in lung cancer cell lines HKULC2, HCC827, HKULC4, H23, HCC1833, H1299 and H1650. Dead cells, cell debris and doublets were gated out. Compensation for background fluorescence was performed by measuring target signals of single color controls and negative controls. Data were presented in 2D diagrams plotting either PE or FITC signals against an irrelevant channel ECD® (also known as PE-Texas Red). The average percentages from 3 individual analyses were presented in Table 1.
Figure 2.
In vitro tumorigenecity of NSCLC cells.
A. Representative immunoblot of H1650, HKULC2, H1299, HKULC4, HCC827, H23, HCC1833 and PLC8024 total protein lysate using antibodies against CD44 and CD133. Actin was used as a loading control. The blot was representative of three individual experiments. B. Overall view to show density of SB formation from CD44+ and CD44− H1299 cells after culturing in RPMI medium with EGF, FGF and insulin supplement for 21 days (upper panel), and morphology of SB at weekly intervals (middle panel). Cells of the first generation SB were disaggregated and analyzed by flow cytometry (lower panel). The CD44+ to CD44− (81.58% to 9.68%) ratio was similar to that of the parental H1299 cells (81.3% to 18.7%, Table 1). SSC, Side Scatter Channel. C. Representative fields of colony formation of unsorted, CD44+ and CD44− H1299 cells in soft-agar after culturing for 21 days. Insets show magnified views of representative colonies from the respective cells taken under the same magnification. CD44− subgroup showed significantly fewer colonies compared to both CD44+ and unsorted subgroups. (**, p<0.01). Histograms represented average of 3 independent observations from separate experiments.
Table 1.
Expression of Putative Cancer Stem Cell Markers in NSCLC Cell Lines.
Figure 3.
In vivo tumorigenicity assay of H1299 cells.
A. Representative pictures of primary xenograft tumors. Tumor initiation was achieved by 200,000 unsorted cells (3/4 mice). From CD44+ cells, tumors were initiated by 10,000 (1/4 mice), 50,000 (3/4 mice) and 100,000 cells (4/4 mice). No tumor was formed with 200,000 CD44− cells (0/4 mice). PBS, saline injection control. B. Flow cytometry analysis of disaggregated primary and secondary xenografts showed similar CD44+ to CD44− subgroup ratios as parental H1299 cells. X-axis: CD44 expression, FITC channel; Y-axis: Dead cells stained by PI, PE channel. C. Semi-quantitative RT-PCR analysis of primary xenografts showed expression of CD44 standard form (CD44s) and pluripotency genes (POU5F1, NANOG, SOX2) in CD44+ but not CD44− subgroups. D. IHC analysis of xenografts from sequential generations of recipient mice showed CD44 expression in cell membrane distribution. Host cells of mouse origin present in the tumor stroma were not stained. Tumor growth curves of the primary, secondary and tertiary xenograft tumors formed by unsorted and CD44-sorted H1299 cells were plotted by observing and measuring tumor formation weekly for up to 9 weeks, afterwhich mice were scarified to avoid overgrowth of tumors.
Table 2.
In Vivo Tumorigenecity of Marker+/− NSCLC cells.
Table 3.
In vivo Serial Transplantation Experiments of CD44+ and CD44− Cells Sorted from H1299 Xenograft Tumor.
Figure 4.
Functional characterization of CD44-selected H1299 cells.
A. Immunofluorescence characterization of SB from CD44+ subpopulation, showing co-expression of OCT4, NANOG and SOX2. DAPI, control; Green, OCT4-FITC; Red, SOX2-Texas Red; Blue, pseudocolor of NANOG by PE-conjugated antibody. B. Semi-quantitative RT-PCR analysis showed pluripotency genes, SNAI1, CDH2, VIM, CD44s and CD44 v3,5,6,10 expression in CD44+ cells. CD44− H1299 cells lack these expression except for CDH2 and VIM. C. CD44+ cells were more resistant to the apoptotic effects of 5 µM cisplatin compared to CD44− in H1299 and H1650 cells after 24 hr incubation. Histograms represented average of three individual experiments of unsorted, CD44+ and CD44− cells before and after cisplatin treatment.
Figure 5.
IHC analysis of human lung tissues.
A. CD44 expression in pseudoglandular stage of embryonic lung (left) and regenerating pneumocytes of lung with diffuse alveolar injury (right). Inset shows CD44+ basal cells in normal bronchial epithelium. B. Representative cases of adenocarcinomas (AD) showing moderate (left) and strong (right) CD44 expression in cell membrane distribution. In some cases of squamous cell carcinomas (SCC), CD44 expression was particularly prominent in tumor cells at the periphery (arrow) than central regions (asterisks) of tumor islands which are generally regarded as stem cell sites. Scattered small lymphocytes and marcophages also showed CD44 expression in the tumor stroma. C. Kaplain Meier survival analysis of patients with AD (upper panel) and SCC (lower panel) stratified according to CD44+ (tumors with moderate/strong expression) or CD44− (absent/weak, focal staining). PFS, progression-free survival in months; OS, ovrall survival in months.